Add to Calendar
2019-07-09 15:00:00
2019-07-09 16:00:00
America/New_York
The Sparse Tensor Algebra Compiler
Abstract: Tensor and Linear Algebra are powerful tools with applications in data analytics, machine learning, science, and engineering. The massive growth of data in these applications makes performance critical. For applications that use sparse tensors, where most components are zeros, programmers must choose between libraries with hand-optimized implementations of select operations and generalized software systems with poor performance. In this talk, I will present compiler abstractions and techniques that combine tensor expressions with specifications of sparse irregular tensor data structures to produce efficient parallel source code. I will show solutions to the three main problems of sparse tensor algebra compilation: how to represent tensor data structures, how to characterize sparse iteration spaces, and how to generate code to coiterate over irregular data structures. I will also show how to optimize sparse tensor algebra code in a compiler and how to programmatically map sparse data to tensors. We have implemented these techniques in the TACO sparse tensor algebra compiler. It is the first compiler to generate sparse code for any basic tensor expression on many sparse tensor representations. The generated code matches or exceeds the performance of hand-optimized libraries while generalizing to any expression and many user-specified irregular data structures. Bio: Fredrik Kjolstad is a PhD student at MIT, working with Saman Amarasinghe on topics in compilers and programming languages. He will join Stanford as an Assistant Professor in 2020. He received his master degree from the University of Illinois at Urbana-Champaign and his bachelor degree from the Norwegian University of Science and Technology in Gjøvik. He has received the Eureka and Rosing prizes for his bachelor project, the Adobe Fellowship, a best poster award, and two best paper awards.Livestream: https://www.youtube.com/channel/UCYs2iUgksAhgoidZwEAimmg/live
G32-449 (Patil/Kiva)
July 09
June 11
Add to Calendar
2019-06-11 15:00:00
2019-06-11 16:00:00
America/New_York
The Resurgence of Software Performance Engineering
Livestream Link: https://www.youtube.com/channel/UCYs2iUgksAhgoidZwEAimmg/liveAbstract: Today, most application developers write code without much regard for how quickly it will run. Moreover, once the code is written, it is rare for it to be reengineered to run faster. But two technology trends of historic proportions are instigating a resurgence in software performance engineering, the art of making code run fast. The first is the emergence of cloud computing, where the economics of renting computation, as opposed to buying it, heightens the utility of application speed. The second is the end of Moore's Law, the 50-year technology trend which has, until recently, relentlessly doubled the number of transistors on a semiconductor chip every two years. With the attenuation of this major source of computing performance, application programmers will increasingly find themselves turning to software performance engineering in order to develop innovative products and applications. Bio: Charles E. Leiserson received his B.S. from Yale University in 1975 and his Ph.D. from Carnegie Mellon University in 1981. He joined the faculty of the Massachusetts Institute of Technology in 1981, where he is now the Edwin Sibley Webster Professor in MIT's Electrical Engineering and Computer Science (EECS) Department. He is Associate Director and Chief Operating Officer of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL), the largest on-campus laboratory at MIT, where he also leads the Supertech research group. He is a Margaret MacVicar Faculty Fellow, the highest recognition at MIT for undergraduate teaching. He is a Fellow of four professional societies — AAAS, ACM, IEEE, and SIAM — and he is a member of the National Academy of Engineering. He has received many Best Paper awards at prestigious conferences, as well as major awards, including the ACM-IEEE Computer Society Ken Kennedy Award, the IEEE Computer Society Taylor L. Booth Education Award, ACM Paris Kanellakis Theory and Practice Award, and the ACM and Hertz Foundation Doctoral Dissertation Awards.
G32-449 (Patil/Kiva)